During a period of 13 months, 1273 consecutive samples of water received at this laboratory were submitted to a modified presumptive coliform test which included single tubes of four different media. These were: (1) MacConkey broth, (2) formate lactose glutamate medium (pH 7·5), (3) lactose glutamate medium (pH 7·5), and (4) Folpmers's glucose glutamic acid (pH 6·0). 955 samples gave exactly the same results in all four media, but the remaining 318 samples produced differences which enabled comparisons to be made between the media.

As compared with MacConkey broth, the three glutamic-acid media produced between 24–30% more isolations of Esch. coli, and the formate and the glucose media produced at least the same total number of true coliform organisms (including Esch. coli) with appreciably fewer false positive reactions. The lactose glutamate medium gave no false positive reactions but, through the suppression of coliform organisms other than Esch. coli, reduced by 16% the total coliform yield. MacConkey broth gave the largest early (18 and 24 hr.) yield of positive reactions, but the results at the end of 24 hr. showed the formate medium to be not far behind MacConkey broth and appreciably ahead of the glucose medium.

Further experiments with the formate lactose glutamate medium adjusted to different pH's, ranging from 6·0–7·5, indicated that a medium of pH 6·7 provided optimal conditions for the early development of both acid and gas. It was therefore decided to test this observation more fully with routine samples. During a period of 4 months, 279 unselected consecutive samples of water (of which 78 gave positive results) and 57 consecutive unchlorinated water samples (56 of which gave positive results) were examined by a multiple-tube technique in both MacConkey broth and formate lactose glutamate medium (pH 6·7). In the former series the MacConkey broth contained the inhibitory bile salt previously used; in the latter series this was replaced by a relatively non-inhibitory bile salt. In both series the formate medium yielded an appreciably increased total of coliform organisms (12 and 17%) including an increased total of Esch. coli (27% in the former and 13% in the latter series), and in both series also the results at 18 hr. were abreast with, and at 24 hr. ahead of, those obtained in MacConkey broth.

With certain reservations, it is considered that formate lactose glutamate medium (pH 6·7) can be offered as a suitable alternative to MacConkey broth for the presumptive coliform test of water. Wider trials of this medium would be required before this claim could be fully substantiated.

My thanks are gratefully accorded to Drs W. H. H. Jebb and A. H. Tomlinson of the Oxford Public Health Laboratory for invaluable advice during the investigation and to Dr Ian Sutherland of the M.R.C. Statistical Research Unit for statistic appraisal of the results. I am also greatly indebted to my technicians Mr J. H. Evans, F.I.M.L.T., and Mr G. H. Lowe, F.I.M.L.T., for their unstinting help.

1. The mesenteric lymph nodes and faeces of pigs, cattle, sheep, dogs and cats and the intestinal wall and other organs of chickens in Essex, England, have been examined for salmonellae. These animals were either healthy or were not suffering from clinical salmonella infection.

2. Salmonellae were isolated from the mesenteric lymph nodes of sixty (12%) of 500 pigs, nine (4·5%) of 200 dogs, five (2·5%) of 200 cats and none of 200 cattle and 100 sheep. None was found in the chickens.

3. One (0·5%) of the cats and one (0·5%) of the dogs and six (1·2%) of an additional 500 healthy pigs were excreting salmonellae in their faeces.

4. Of the seventeen serotypes found in pigs, Salm. typhi-murium, Salm. anatum and Salm. cholerae-suis occurred most frequently.

5. All strains of Salm. cholerae-suis were isolated in brilliant green MacConkey broth; selenite broth was unsuitable for this purpose.

6. Despite the high isolation rate from pigs, clinical salmonella infection was diagnosed relatively infrequently at this laboratory. Salm. cholerae-suis was isolated from only one of seventeen epidemiologically unrelated cases of necrotic enteritis in pigs.

7. The results are discussed from the public health and agricultural view points with particular regard to the part infected feedingstuffs may play in causing salmonella infection in animals.

The expenses of the investigation were, in part, defrayed by grants from the Agricultural Research Council, whose help I gratefully acknowledge.

I am grateful to Mr A. J. Harman and Miss Linda Kelly for their capable technical help and to Dr K. C. Sellers for useful suggestions made during the course of this work. I am also indebted to Dr Joan Taylor for typing all the salmonella cultures and to Mr L. G. Smart for collecting the lymph nodes from the cattle, sheep and pigs.

Re-examination of the U-virus of Philipson & Wesslén has shown it to be a typical ECHO virus closely related to Type 11 of this group. Ultrafiltration and ultra-centrifugation techniques indicate that the virus particle has a diameter of about 27 mν and serological studies confirm its antigenic relationship to the ECHO group.

The virus has been successfully transmitted via the nasal route to volunteer adult subjects. All of them became infected but the clinical picture showed no more than a very mild gastro-intestinal illness or a slight disturbance of the general well-being.

Passage in monkey kidney cells appeared to attenuate significantly when compared with short-term passage in human embryo-lung culture, and this seemed to affect the isolation rate from subjects so infected.

The virus is highly immunogenic. Antibodies to the U-virus have been found in normal persons residing in Great Britain.

We wish to thank the volunteers for their willing co-operation and Miss J. B. Macdonald for help with the clinical observations. Dr A. D. Macrae supplied the ECHO Type 11 prototype virus and immune serum and Prof. C. H. Stuart-Harris supplied the pooled human sera. The Director, Microbiological Research Establishment, Porton, afforded the use of the ultracentrifuge. Monkey kidneys were supplied by the M.R.C. polio vaccine control laboratories. Mr M. C. Brebner and Miss P. K. Pearce assisted with the laboratory work.

The Mahoney, Type 1, poliomyelitis virus had been through six to-and-fro monkey kidney-tissue culture and suckling mouse passages. Early paralysis was produced in groups of mice inoculated directly from infected tissue-culture fluid, but mouse-to-mouse passage failed to produce paralysis except in an occasional mouse. At the 7th tissue-culture mouse passage, the virus harvested from paralysed mice was inoculated into 100 mice. One mouse developed paralysis within 24 hr. From this mouse a variant of Mahoney virus was isolated which has now been through seventy-five mouse-to-mouse passages and produces paralysis in 90–100% of mice within 3 days.

The authors would like to express their deep appreciation of Prof. A. Kipps' constant interest and encouragement.

1. The recovery, after poisoning, of the rat population of eight sewer systems was measured by the bait-census method.

2. In three of the systems the mean rate of increase was of the order of 3% per week. This is attributed solely to breeding and is regarded as typical for rats in sewers.

3. Higher rates of increase (up to 11·8%) were also recorded but are believed to have been partly due to immigration or to have been only apparent (as a result of census difficulties).

4. The consequences for control of a rate of increase as low as 3% per week are discussed.

Our thanks are due to eight Local Authorities, their Chief Public Health Officers and members of their staffs, without whose help the work would have been impossible, and to a number of colleagues, especially Mr J. H. Greaves, who helped with much of the census work.

The form of the killing curve obtained by bombaring micro-organisms with high energy electrons enables an estimate to be made of the numbers of individual viable organisms present in the aggregates or clumps comprising the sample. Samples of Staph. aureus collected from the air of two hospital wards have been found to consist of aggregates containing, on the average, only about four viable cocci per air-borne particle. These samples were taken during quiet periods and during periods of active dispersion of the organism but there was no active spread of staphylococcal disease at any time.

The work on the linear accelerator was carried out with the aid of grants provided by the Treasurer and the Board of Governors from the Discretionary Fund of St Bartholomew's Hospital, to whom we wish to express our gratitude.

1. Neonatal serum proteins and those of maternal colostrum have been examined in the Spinco analytical ultracentrifuge and the Perkin-Elmer electrophoresis apparatus. To facilitate correlations between the two types of result, certain protein fractions have been prepared electrophoretically and examined in the ultracentrifuge.

2. The electrophoretically distinguishable components of precolostral serum, albumin, α and β globulin and traces of γ globulin sedimented mainly as a single boundary which showed evidence of a slower component (probably fetuin) with a sedimentation constant of about 3S. A small proportion of a macroglobulin component (S020 ≅ 16S), associated with the α globulin, was also present. In conformity with the very low γ globulin, there was a complete lack of globulin with S020 ≅ 6·5–7·0S. The β globulin and the bulk of the α globulin sedimented closely with albumin, whose sedimentation constant in an electrophoretically isolated fraction was somewhat lower than normal, probably as a result of α-globulin contamination.

3. The electrophoretic and ultracentrifuge analysis of postcolostral calf serum showed evidence of passive absorption from the maternal colostrum of immune lactoglobulin. There was a rise in the concentration of total serum protein of which up to 51·6% was immune lactoglobulin, of sedimentation constant 6–6·5S. No such component occurs in precolostral sera. The electrophoretic and ultracen-trifugal characteristics of this globulin were not significantly altered as the result of absorption. The acquisition of immune lactoglobulin was accompanied by a simultaneous fall in serum albumin concentration.

4. Examination of colostrum, rennin-produced colostral whey, and of electrophoretic fractions prepared from colostrum showed the presence of a small proportion of a macroglobulin (S020 ≅ 18S), a well-defined immune lactoglobulin of S020 ≅ 6·4S and electrophoretic properties in the γ-range, and a group of electrophoretically more mobile components with sedimentation constants ranging downwards from 3S.

5. Electrophoretically prepared fractions from the serum of a deprived calf 27 days old showed sedimentation properties in conformity with the results on precolostral sera, with the macroglobulin component being again associated with α globulin. The remainder of α and β globulin sedimented alongside albumin and the additional γ-globulin component in a single boundary of S020 ≅ 6·5–7S.

6. The concentration of autogenous γ globulin developed by calves which had been deprived of colostrum increased with time and was generally in fair agreement with that of the 6·5–7S component.

The authors wish to thank Sir Alan Drury, F.R.S., for his interest and encouragement during the course of this work, Mr N. Buttress, Mr D. Hardman and Miss J. Mallon for technical assistance and Mr J. Clark for assistance in the care and handling of the calves.

Good fixation of complement was obtained with the infective virus fraction of the Mahoney Type 1 poliomyelitis virus which had been adapted to suckling mice. The presence of a soluble antigen, specific to Type 1 virus, was also demonstrated but was small in amount.

The infective virus fraction of the suckling mouse adapted MEF1 Type 2 poliomyelitis virus produced as good fixation of complement as did Type 1, but the fixation of complement by the MEF1 soluble antigen was far greater than that obtained with the Mahoney strain.

The author would like to thank Prof. A. Kipps for his interest and encouragement and Miss M. Butchart for her valuable technical assistance. She is also indebted to Dr H. Malherbe of the Poliomyelitis Research Foundation, Johannesburg, for the Type 1 monkey immune serum.

Colorimeters and non-recording spectrophotometers for visible and ultra-violet light have been used for many years and are to be found in most microbiological laboratories. With the need for greater speed of operation, recording instruments are now coming into more general use. During the last two decades recording infra-red spectrophotometers have been developed and these have enabled the absorption measurements on micro-organisms to be extended into the infra-red region of the spectrum. Two factors have tended to retard the use of infra-red spectrophotometry. One is the high initial cost of the equipment and the other is the large absorption of infra-red radiation by water. Despite these difficulties, a considerable amount of work has now been done and it seems profitable to review the varied applications to which infra-red spectroscopy has already been put and to indicate the results which have been obtained by its use.

1. A new phage-typing scheme for Salm. typhi-murium is described which initially defines thirty-four types of the organism.

2. Types recognizable with the older scheme of Felix and Callow remain distinct with the new system. However, the latter method offers a finer subdivision of the organism than does the former.

3. Thirteen of the twenty-nine typing phages used in the new scheme were prepared by adaptation of Salm. paratyphi B typing phage 3b, and nineteen are serologically indistinguishable from this phage.

4. The lysogenic properties of the new types are discussed. Twenty-three out of the thirty-three type strains yielded both thermolabile and thermostable phages and a further three yielded thermolabile phages only. No phages could be isolated from the remainder.

5. The distinctive host ranges of the phages carried by most types can be used for type identification.

I am indebted to Miss Alison M. Dykes for her valuable assistance with this work; to Mr F. J. Flynn for help in the preparation of phage stocks; and to Dr E. S. Anderson for many helpful discussions and advice in the preparation of this paper.

A total of 454 sampling experiments were performed on the groupings of 100 samples each from separate Poisson populations in order to establish estimates of frequency of rejection of the hypotheses of homogeneity by means of the z, or index of dispersion, test. Results of these artificial realizations are presented in Figs. 1–6, and compared with the corresponding non-central ξ2-distributions.

It is a pleasure to mention the technical assistance of Dennis Hamilton of the Research Computer Laboratory in programming the detailed computational phases of this investigation, which was partially supported by the State of Washington Initiative 171 Funds for Research in Biology and Medicine.